use File::Find;

find(\&wanted, '/usr/local/src');

sub wanted {
    open(FL, $File::Find::name);
    # process stuff here #
    close(FL);
}
[download]

broquaint

sub handle_dir
{
  my $d = shift;
  my $sd = shift || '';
  my @all = ();

  opendir (DIR, $d) or die "Broken!";
  while (defined(my $file = readdir(DIR))) { push @all, $file; }

  chdir ($d);

  foreach my $file (@all)
  {
    next if $file =~ /^\.$/;
    next if $file =~ /^\.\.$/;
    my $fp = $d . "\\" . $file;

    if ($file =~ /htm$/i) { &handle_html ($fp); }
    if (-d $file) { &handle_dir ($fp); }

    chdir($d);
  }

  chdir("..");
}
[download]

You're much better off using File::Find than reinventing the wheel like this.

Firstly, your code will only work on Windows, or any other OS where the directory separator is \. File::Find deals with this transparently. Also you're not checking for failure. What happens if the chdir ($d) fails, maybe because the directory specified in $d doesn't exist?

The first two lines to match for the current or parent directory could be rewritten as next if $file =~ /^\.{1,2}$/;.

The first two lines to match for the current or parent directory could be rewritten as next if $file =~ /^\.{1,2}$/;.

But it would be better not to, because it is a potential time-bomb waiting to explode. Why do people keep dragging out this broken regexp, when a couple of conditionals will do the trick?

next if $file eq '.' or $file eq '..';

--

g r i n d e r

hokay - well ..
Initiates rush in where abbots fear to tread I guess
in future, for sake of any reputation I wanna maintain, will pause before submitting (:
thems the breaks
was a snippet from a program to generate an index of a cd of html content ..
I mean, it works, but s'not exactly value-added-perl
live n learn etc
the 'qif;